NVIDIA releases AI-Q Blueprint for connecting AI agents

Blueprint promotes the dual improvement of efficiency and effectiveness in business operations.
AI agents, as a new type of digital workforce, are changing business operation models, automating complex tasks, and unleashing new efficiencies. Nowadays, with the help of collaborative capabilities, these intelligent agents can work together to solve complex problems and even have broader impacts.

Enterprise users across industries such as sports and finance can now accelerate the acquisition of intelligent agent advantages through the new NVIDIA Blueprint - AI-Q. AI-Q is used to develop agent-based systems with reasoning capabilities to extract knowledge from enterprise data.
Utilizing NVIDIA AI-Q and AgenTIQ toolkits
Building a smarter proxy AI system
AI-Q provides a reference for integrating NVIDIA accelerated computing, partner storage platforms, and software and tools (including the new NVIDIA Llama Nemotron inference model). AI-Q also provides a strong foundation for enterprises to build a digital workforce that can break through proxy silos and accurately and quickly handle complex tasks.
AI-Q quickly integrates multimodal extraction and excellent retrieval capabilities through NVIDIA NeMo Retriever, NVIDIA NIM microservices, and AI agents.
This blueprint is supported by the new NVIDIA AgenTIQ toolkit, which enables seamless heterogeneous connectivity between agents, tools, and data. As an open-source software repository, AgenTIQ has been released on GitHub for connecting, analyzing, and optimizing AI agent teams driven by enterprise data to create multi-agent end-to-end systems. It can be easily integrated with existing multi-agent systems, either partially or as a complete solution, with a simple process of 100% opt in.
The AgentiQ toolkit also enhances transparency through complete system traceability and performance analysis, enabling organizations to monitor performance, identify inefficiencies, and gain a detailed understanding of how business intelligence is generated. This analysis data can be combined with NVIDIA NIM and NVIDIA Dynamo open-source libraries to optimize the performance of proxy systems.
New enterprise AI intelligent agent workforce
As AI agents become digital employees, IT teams will provide support for onboarding training. The AI-Q blueprint and AgenTIQ toolkit support digital employees by enabling collaboration between intelligent agents and optimizing performance across agent-based frameworks.
Enterprises using these tools will be able to more easily connect AI agent teams across solutions, such as Salesforce's Agentforce, Confluence, and Jira's Atlassian Rovo, as well as the ServiceNow AI platform for business transformation, to break down silos, simplify tasks, and shorten response times from days to hours.
AgentIQ also integrates frameworks and tools such as CrewAI, LangGraph, Llama Stack, Microsoft Azure AI Agent Service, Letta, etc., enabling developers to work in their preferred environment.
Azure AI Agent Service has integrated AgentIQ, which enables more efficient orchestration of AI agents and multi-agent frameworks using AgentIQ's fully supported Semantic Kernel.
Many industries are integrating visual perception and interaction functions into intelligent agents and copilots.
Financial services leader Visa is using AI agents to simplify network security and automate large-scale analysis of phishing emails. By utilizing AI-Q's analyzer capabilities, Visa can optimize the performance and cost of intelligent agents, maximizing the efficient role of AI in threat response.
Start using AI-Q and AgentiQ immediately
NVIDIA Metropolis VSS blueprint integrates AI-Q, which enables multimodal agents to combine visual perception with speech, translation, and data analysis to enhance intelligence.
A proxy system built using AI-Q requires a powerful AI data platform. NVIDIA's partners are providing customized platforms for continuous data processing, enabling AI agents to quickly access knowledge for inference and respond to complex query requests.